The structural details of Golgi membranes and their associated vesicles are still undefined. Both the continuity between adjacent Golgi elements and the relationship between the trans-Golgi cisternae and theie associated network of vesicles and tubules (the Trans-Golgi Network, or TGN) are unresolved. We are approaching this problem with two 3-D technologies: (1) selective stains and tomography and (2) antibodies or enzymatic markers, in conjunction with thin or super-thin serial sections. Ten tomograms have been prepared from chemically fixed cells (Ladinsky et al. 1994) and 5 from fast-frozen, freeze-substituted cells. These tomograms reveal clathrin coated vesicles and other vesicles with a novel coat morphology forming in the TGN. We are now focusing our attention on cells fixed by rapid freezing and freeze-substitution to get information about membrane arrangements in cells preserved by high quality methods. Preliminary work with these specimens confirms our identification of different coat morphologies on vesicles budding from the TGN and suggests unexpected relationships among the successive cisternae in the Golgi stack. Issues of membrane continuity between stacks and between Golgi elements and endosomes are now being examined in this well preserved material. Several novel features of Golgi morphology have emerged, such as fenestrae of various sizes on many of the cisternae and finger-like tubules that project from an interior cisterna through a fenestra in an adjacent cisterna, into the TGN or CGN. We have also found cis and/or trans cisternae that connect across noncompact regions of the Golgi (See Research Highlight #1).